[goatherders]

I don't much care about the semantics of product market fit, but there are a lot of companies - some with astronomical valuations - that sell things at prices that won't sustain the business. These guys could have taken other people's money and done the same thing. "We just need to buy market share until the autonomous moving robots are available!"

I applaud them for not doing so. And FWIW I usually find these autopsies annoying but this one is well written and insightful. Applause.

[thisisit]

This might sound cynical but given the way the conclusion has been phrased, it is most likely they were on path to "autonomous robots" but no VC was interested in their unit economics.

Sure, there are companies with astronomical valuations with unsustainable business, but for whatever it is worth they seem to be good at marketing/presenting to VCs too.

[underwoodley]

Moving awkwardly shaped pieces of furniture up and down stairs without damaging them or the building is one of the jobs that is not likely to be performed by robots for a long time to come. It is very difficult to automate, and very easy for humans to do.

[Mtinie]

> It is very difficult to automate, and very easy for humans to do.

Easier, but if you’ve spent time attempting to do it, you’ll agree it is not a trivial type of task by any means.

Judging the center of balance for awkwardly shaped furniture, while safely navigating the interior of a multi-level residence is not on my short list of “easy” tasks I’d sign up for on a regular basis.

In many ways, it’s a task that should benefit from automation in the future. Endurance, strength, precision, and the ability to precalculate the best available three-dimensional routes required to move a piece of furniture are all, individually, solved problems. Bringing them together with a form factor that can adapt to different environments — flooring material, stairs, etc. — and at a price point that encourages use is the challenge.

[underwoodley]

The price point is exactly the challenge. Robots to do this on the Moon, or inside nuclear power stations after an accident, will make economic sense quite soon. Robots to do this competing with a semi-skilled worker who may have substance abuse problems, a criminal record, or be undocumented (all factors which make them cheaper to employ) are not going to be seen for some time. Most desk jobs will be made obsolete before this is an automation priority.

[notyourday]

Companies that do it successfully for people who do not have replaceable furniture sometimes rent small cranes in cities like NYC.